OpenAI Missed How Much Users Loved GPT-4o and Its Personalities
347OpenAI just launched GPT-5, but it seems the company has learned a tough lesson. They realized they might be out of touch with what their users really want. After the release, many fans of the old GPT-4o model were surprised and upset when OpenAI deprecated it. They loved GPT-4o’s unique personality and the way it interacted. Some users even demanded its return. OpenAI listened and brought GPT-4o back for paying subscribers, but it’s clear they didn’t see just how passionate some users were about that specific model and its “personalities.”
In a recent podcast interview, OpenAI’s longtime ChatGPT boss, Nick Turley, talked about the backlash, mostly seen on Reddit. He said the comments showed how divided users are. Some really loved GPT-4o, while others preferred GPT-5. Turley found it interesting how strongly people felt about their choices. This made him step back and rethink how users see and use these models. It was a wake-up call for the team.
Turley admitted he was initially confused about why users felt so strongly. Right after ChatGPT launched, he hired a data scientist to help figure out what people liked. Turley explained that during late 2022 and early 2023, he was hearing all sorts of different stories from users about why they loved ChatGPT. Some praised its ability to write, code, chat, or answer questions. But those reasons only scratch the surface of what makes ChatGPT so popular.
He said that over time, he learned the main ways people use ChatGPT—like writing, coding, casual chatting, and searching for information. Still, he pointed out that the way people use AI is much more complicated than just those four categories. Many users rely on ChatGPT for emotional support, as a sort of therapist or life coach, which isn’t as obvious from the outside. That’s a big part of why some users become so attached to specific models and personalities.
Turley also mentioned that, although some things have changed over the past year, ChatGPT is still mostly the same. Its core use cases remain the same—writing, coding, chatting, and searching. The improvements made recently have mostly been about tweaking the model’s behavior, personality, and how likely it is to refuse certain requests. He admits he’s still figuring out what’s happening under the hood.
Interestingly, OpenAI’s CEO, Sam Altman, seems to understand the situation better than Turley. After the initial backlash over GPT-4o, Altman tweeted that OpenAI has been paying close attention to how people become attached to specific models. He said they’ve been tracking this for over a year, especially as many start using ChatGPT as a therapist or life coach.
Turley acknowledged that OpenAI has been monitoring these trends for a while but seemed only recently to be asking why people feel so strongly about ChatGPT. That gap in understanding might be a real problem. If the company doesn’t fully grasp why users love certain models or personalities, it could affect future updates and user satisfaction.
In the end, it’s clear that OpenAI might need to rethink how they approach user attachment and model personalities. The passion people have for specific versions shows that AI isn’t just about functionality anymore; it’s about emotional connection. As AI becomes more embedded in daily life, understanding those emotional ties will be crucial for building better, more accepted tools.
What do you think?
It is nice to know your opinion. Leave a comment.